You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by "Varun Thacker (JIRA)" <ji...@apache.org> on 2018/01/27 22:57:00 UTC
[jira] [Commented] (SOLR-11920) Differential file copy for
IndexFetcher
[ https://issues.apache.org/jira/browse/SOLR-11920?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16342360#comment-16342360 ]
Varun Thacker commented on SOLR-11920:
--------------------------------------
+1 to the idea.
Restore Core does a local copy instead of the hard link approach that you're proposing in this patch. I haven't looked up if hard link is the right approach but if it is we could do something similar in restore as well.
[https://github.com/apache/lucene-solr/blob/master/solr/core/src/java/org/apache/solr/handler/RestoreCore.java#L96]
Maybe we should turn the two INFO log lines into DEBUG? I know they are currently INFO as well but seems like we should change it
# LOG.info("Don't need to download this ...
# {{LOG.info("Skipping download for ... }}
> Differential file copy for IndexFetcher
> ---------------------------------------
>
> Key: SOLR-11920
> URL: https://issues.apache.org/jira/browse/SOLR-11920
> Project: Solr
> Issue Type: Improvement
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Ishan Chattopadhyaya
> Assignee: Ishan Chattopadhyaya
> Priority: Major
> Attachments: SOLR-11920.patch, SOLR-11920.patch
>
>
> In the case of fullCopy=true, all files are copied over from the leader/master irrespective of whether or not that exact file exists with the replica/slave. This is wasteful, esp. in tlog replicas or pull replicas, when only a fraction of the total files are different.
> This stems from SOLR-11815.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org